Effects of density-dependence, environment and species interaction during spawning and incubation on population dynamics of pink and sockeye salmon in the Auke Lake system, southeast Alaska

Abstract

Mechanisms that regulate or influence fry and smolt production of pink and sockeye salmon in the Auke Lake system, southeast Alaska, were studied with special emphases on effects of: (1) density-dependence; (2) environmental effects; and (3) species interaction. There has been an increasing trend in the residuals of spawner-recruit models for pink and sockeye salmon since the late 1970's. A strong positive influence of precipitation was found in sockeye smolt production. Estimated spawner capacity of pink salmon was 15 times greater than sockeye salmon in the system. Pink salmon spawners varied in stream life (5-11 days), spawning efficiency (30-70% of daily cohorts of females retained less than 500 eggs at death), and habitat selection (30-70% spawned in Auke Creek rather than Lake Creek, the inlet stream). Variation of these attributes in female pink salmon was explained by various environmental variables using generalized linear models. Repeated use of limited spawning grounds by Pacific salmon, i.e., redd superimposition, can cause density-dependent mortality. Pink salmon egg loss from part of Auke Creek, estimated by a series of mark and recapture experiments, was roughly proportional to spawner abundance and not related to discharge. The maximum daily egg loss was estimated to be 3-400,000 eggs. Eggs in samples were more advanced in development later in the season. Eggs were washed out from the streambed due to redd superimposition; eggs spawned by early pink salmon spawners suffered higher mortality than eggs spawned by later spawners. The peak sockeye spawning preceded the peak pink spawning by approximately one week in Lake Creek, and the major spawning areas of sockeye salmon occurred approximately 250-350 m upstream from those of pink salmon. Microhabitat selection measured by four variables differed significantly between the species, but discrimination between the species was impossible because of large overlaps. Habitat variation was greater among different runs of sockeye salmon than between the two species. Sockeye salmon shifted spawning sites from riffles to pools as the season progressed.